Usually you just draw a 2D or 3D picture and say "n" while pointing to it. e.g. I had a professor that drew a 2D picture on a board where he labeled one axis R^m and the other R^n and then drew a "graph" when discussing something like the implicit function theorem. One takeaway of a lot of linear algebra is that doing this is more-or-less correct (and then functional analysis tells you this is still kind-of correct-ish even in infinite dimensions). Actually SVD tells you in some sense that if you look at things in the right way, the "heart" of a linear map is just 1D multiplication acting independently along different axes, so you don't need to consider all n dimensions at once.